The requirements outlined in International Convention for the Safety of Life at Sea, 1974 (SOLAS) for the design of lifeboats and lifeboat release mechanisms focus primarily on the engineering design requirements for safe operation.4 The Pacmonarch's lifeboat fully complied with all SOLAS requirements. A human factors or ergonomic analysis could identify design deficiencies that, instead of being based on engineering principles, are based on how the design influences the effectiveness, efficiency, and safety of seafarers using this type of evacuation system. When considered in isolation, a design aspect may not contain a serious deficiency. However, when all the lifeboat's components are integrated and considered as a whole, they present a risk whereby the lifeboat may be raised or lowered with the hooks improperly reset and the interlock not engaged. This condition could, and did, result in the lifeboat's accidental release and the consequent injury, loss of life, and damage to property.2.0 Analysis The requirements outlined in International Convention for the Safety of Life at Sea, 1974 (SOLAS) for the design of lifeboats and lifeboat release mechanisms focus primarily on the engineering design requirements for safe operation.4 The Pacmonarch's lifeboat fully complied with all SOLAS requirements. A human factors or ergonomic analysis could identify design deficiencies that, instead of being based on engineering principles, are based on how the design influences the effectiveness, efficiency, and safety of seafarers using this type of evacuation system. When considered in isolation, a design aspect may not contain a serious deficiency. However, when all the lifeboat's components are integrated and considered as a whole, they present a risk whereby the lifeboat may be raised or lowered with the hooks improperly reset and the interlock not engaged. This condition could, and did, result in the lifeboat's accidental release and the consequent injury, loss of life, and damage to property. 2.1 Hook Design and Orientation and Ergonomics of Hook Resetting The forward and after hook assemblies both face aft so that the lifeboats can be lowered into the water while the ship is making headway. This could cause the lifeboat to be towed along by its falls, changing the relative angle between the falls and the hook. Orienting the hook openings in this way removes the possibility of the suspension rings slipping out of the hook. However, this arrangement also makes it difficult to see if the after hook has been properly reset. Resetting the hooks is a three-part, two-handed process, in which the right hand has to perform two different actions. The right hand first has to hold down the hook, while the left hand turns the blocking knob to bring up the blocking lever. The right hand then has to release the hook, then reach down and pull the reset lever up to the reset position, so that the lever lines up with the corresponding reference marks. In the forward hook assembly, the reset lever and the red painted reference marks on the cheek plates face into the lifeboat. This orientation makes it easy for the operator to check whether the hook has been correctly reset. In the after hook assembly, the reset lever and the reference marks face aft and away from the operator. This orientation requires the operator to lean half outside the small hatch and to twist around to view the alignment of the reference marks on the reset lever and the cheek plates. This process is awkward, especially since the operator is wearing a bulky lifejacket and a hard hat and if the lifeboat is moving in a seaway. This difficulty reduces the likelihood of the visual check being performed. Additionally, once the operator has visually checked for the correct resetting and has re-entered the lifeboat, s/he is unable to continuously verify that the reset lever has remained in place and has not slipped down. With the forward hook, the check can easily be done at any time. With the after hook, the check can only be done by repeating the procedure described above. Because of the after hook's awkward orientation, the operator tends to use other sources of feedback that, although more readily available, are not reliable indicators of whether the reset lever is in the correct position. In particular, the red alignment line drawn on the side of the hook can be inappropriately used as the main indicator. Thus, the operator looks to see that this red line is aligned and that the hook stays up and feels secure when shaken. The operator then assumes that the mechanism is fully reset and locked. The after hook's aft-facing orientation places its reset lever where the operator cannot easily manipulate it, verify that it has been correctly reset, or see that the lever has remained in place after being reset. 2.2 Interaction Between Gripes, Gripe Release Arrangement, and Lifeboat The following was observed when videos of the test launching of the Pacmonarch's (and of its sister ship, the Pacemperor) starboard lifeboats were analyzed and reviewed in slow motion. Photo19. Still from video showing gripe wrapped around after bollard During the initial part of the launch sequence, the davit's luffing action toward the ship's side causes the lifeboat to swing outward. As the davit draws away from its support frame, it creates an ever increasing gap between itself and the rigid frame. This gap frees the rotating lever, to which the davit end of the gripe is hooked, and allows the lever to turn and release the gripe. The lifeboat's outward motion creates tension in the gripes, which in turn creates angular velocity in the free end of the gripe. The free end of the gripe falls, causing the gripe to loop around the bollard on the lifeboat and, as the relative distance between the free end of the gripe and the bollard reduces, the angular velocity of the free end increases. By the time the davits have reached an angle of about 60, the end of the gripe has gained so much velocity that the gripe wraps itself around the bollard. Since the other end of the gripe is fixed to the deck, the gripe is forced to immediately and violently unwrap itself as the lifeboat continues its swing and the distance between it and the davit frame increases. The gripe's turn around the bollard reduces in diameter and eventually coils around it. At this point, the ring on the free end of the gripe has reached the bollard, and the entire gripe is looped around it. On the day of the occurrence, it is most likely that the end of the after gripe fouled and seized around its bollard as the port lifeboat was lowered. Consequently, the lifeboat took a bow-down attitude, the load partially transferred from the after falls to the gripe, and the after hook separated from its suspension ring. 2.3 Dynamics of the Accidental Separation of the After Hook The entire process of the gripe turning itself up around the bollard takes about two seconds; a complete wrapping and unwrapping takes about four seconds. The precise dynamic behaviour of the gripe is unpredictable and a function of the initial pre-tension in the gripe, the smoothness with which the winch brake is lifted and the lifeboat is launched, the resistance offered by the davits in their hinges, the resistance offered by the rotating lever, the friction between the horns on the davits and the floating blocks, the trim and list of the ship, etc. When the gripe wraps itself tightly around the bollard, the ring at the gripe's end sometimes gets caught between the gripe and the side of the lifeboat. Since the base plate of the after bollard is flat and lies on the flat, vertical stern of the lifeboat, the after gripe end is much more likely to foul in this way than the gripe on the forward bollard, the base of which is radiused backwards to conform to the shape of the lifeboat's bow. Because the gripe is of fixed length and the lifeboat was being lowered away, part of the load from the after falls momentarily transferred to the fouled gripe. The rate of descent of the lifeboat's after end decreased momentarily, but since the falls were still being paid out, the lifeboat's forward end became lower than its after end. The restraint on the after end, caused by the fouled gripe, also held the after end in, closer to the ship side than the forward end. The lifeboat at this point was bow down around 40 (as suggested by matching the witness marks on the damage to the after hook and its suspension ring) when the gripe freed itself and detached from the after bollard. At this angle, the after suspension ring bore against the tip of the hook, and the retaining latch overcame the resistance offered by it and slipped out. Completely separated from the after falls, the lifeboat swung like a pendulum about the forward hook, which was still attached to the forward falls. 2.4 Analysis of the Hook Assembly 2.4.1 Hook Design The reset lever holds the hook in a position where it can bear an applied load. However, this applied load creates turning moments on the hook and the blocking lever and also manifests itself as friction between the blocking and reset levers. The frictional force increases with increasing applied load and is mainly responsible for holding the reset lever in the reset position. If the applied load is removed, so also is the frictional force, and the reset lever is now held in position only by the friction in the teleflex cable and the linkages that connect the reset lever to the central release control assembly. These linkages are well lubricated, and their friction coefficient is low. An analysis of the moments due to the weights of the reset lever arm and the cable connection arm shows that the reset lever has the higher moment.5 Once the suspension rings have been inserted into the hooks, and while the lifeboat itself is riding successive crests and troughs in the waves, this applied load becomes transient and associated with sudden changes in direction of momentum that would tend to allow the reset lever to fall in the opening direction. To prevent such an inadvertent or accidental opening of the hooks, the interlock lever in the central release control assembly has been designed to block the cable connection arms and hold them in position. The reset lever has no other means of ensuring that it stays in place. Figure6. View of hook in load-bearing (reset) condition The hook, the reset lever, and the blocking lever have been designed to rotate about pivots. To swing the blocking lever away and release the hook, the reset lever must first be moved out of the way. Since the reset lever moves in an arc, its contact surface with the blocking lever is also an arc. However, the corresponding contact surface of the blocking lever is planar, thus creating a line contact between these two surfaces (seeSection2.4.2). At its contact surface with the hook, the blocking lever has been given a relief angle of 18. This angle also creates a line contact between them. The retaining latch allows the suspension ring to be inserted into the hook. Because the latch can only open inward, it retains the ring within the hook assembly. To remove the suspension ring from a hook that has been locked by the blocking lever and the reset lever, the retaining latch has to be manually turned inward, creating an opening through which the ring can be slipped. The retaining latch is not designed to take any significant applied loads. 2.4.2 Forces Acting on the Reset Lever The force exerted by the blocking lever on to the reset lever will pass through the centre of rotation of the reset lever, as long as the centre of rotation is in the same line as the centre of its radiused contact surface. If the two centres are not in the same straight line, the force will create a turning moment on the reset lever, tending to move it up or down. The two hook assemblies were tested in a tensile testing machine. A uniformly increasing load was applied to the forward and after hooks at three angular positions of the reset lever (seeAppendixA - Tables1and2). When the reset lever was fully reset, the applied loads created a proportionate turning moment on the reset lever, which tended to move down (toward open). Other angular positions produced a negligible turning moment on the reset lever. The turning moment detected on the hook reset lever indicates that the profile of the lever's load-bearing cam was not of a uniform radius, concentric with the centre of rotation. Because the equipment was new and wear was undetected and unlikely, this profile is a deviation from the designed profile and indicates a manufacturing nonconformity. 2.4.3 The Load-Transmitting Surfaces of the Hooks and their Manufacturing Nonconformities Figure7. The release hook, the frame, and the lever should be replaced when the wear at the contact point between the release hook and the frame reaches 122mm for a 2.5-ton hook. The load-transmitting contact surface on the hook wears with use, reducing the length from the end of its load-transmitting surface to the circumference of its pivot. According to the manufacturer, this minimum distance is 122.0mm. A shorter length can lead to two effects, both potentially detrimental: The contact line between the hook and the blocking lever will shift away from the blocking lever. Since this face of the blocking lever has been given a relief angle of 18, the hook could eventually slip past the blocking lever, rotate, and release the applied load. Although this length on the after hook was found to be around 121.8mm (0.2mm below specification), the design had an adequate factor of safety, and there is no indication that this hook ever slipped past its blocking lever. However, the lifeboat and its hook assemblies were only about a year old, and the lifeboat itself had only been launched about four or five times. This use is not considered sufficient to have caused appreciable wear. As the contact line between the hook and the blocking lever shifts away from the blocking lever, it also allows the hook to change its vertical angle, that is, to revert to its original position. Minimal wear was found on this surface and the original radius was intact. The manufactured length from the end of the hook's load-transmitting surface to the circumference of its pivot was therefore 121.8mm. Because the equipment was new and the wear minimal, this length is a deviation from the designed profile and indicates a manufacturing nonconformity. 2.4.4 Ergonomics of the Alignment Marks for the Reset Lever Photo20. Alignment marks between the cheek plates and the reset lever When the hook is fully and properly reset, the red mark on the tip of the reset lever should line up with the corresponding red mark on the cheek plates of the hook assembly. This alignment provides visual feedback to the operator that the lever is in the correct position and confirms that the hook has been fully reset. If the hook reset indicators are not aligned, the possibility of error exists, since the task now changes from one where a precise judgement is required (whether the red reference marks are aligned) to one where an imprecise judgement is required (whether the reference marks are close enough). Less precise specifications result in more subjective judgements and a greater risk of error. Accurately aligned reference marks are an anchoring aid and prevent a drift in standards. Inaccurately aligned reference marks may lead future operators to accept increasing misalignments as the norm. The misaligned hook reset reference marks did not provide the crew with an accurate indication that the reset lever was in the fully up position and, critically, that the hook was properly reset. 2.5 Central Release Control Assembly Pulling on the release handle moves the cable connection arms upward and pulls the reset levers downward. Teleflex cables connect the cable connection arms to the reset levers on the hook assemblies. As a result, the blocking lever rotates away from the hook, and the hook swings open and releases the suspension ring. The lifeboat is then no longer attached to the falls. The release handle acts on the forward and after cable connection arms simultaneously. To accomplish this, the distance the reset levers have to be moved before releasing the blocking levers is adjusted at the initial setup. The interlock lever prevents the cable connection arms from travelling upward after the hooks have been reset. This lever is mechanically connected to a solenoid coil that is activated by an electrical signal from a hydrostatic pressure switch. When the lifeboat is floating on the water, the solenoid coil moves the interlock away, allowing the operator to pull the release handle and open the hooks. This sequence is known as an off-load release. Figure8. Internal schematic of central release control assembly In case the lowering mechanism malfunctions, which could cause the lifeboat to become suspended from the falls, the hooks are also designed for an on-load release. In this case, the Perspex window in front of the interlock lever's free end is broken, and the interlock is physically lifted off the cable connection arms. The operator can now pull on the release handle to release the hooks and allow the lifeboat to fall to the water. Once the reset levers have been moved into their locking position, they have no intrinsic force to hold them in place. By restricting the movement of the cable connection arms, the interlock lever holds the reset levers up. The interlock lever also holds the hooks in a condition where they can take the applied load and prevents the hooks from accidentally or inadvertently opening. It is therefore critical for the operator to know the position of the interlock lever in relation to the cable connection arms. The two side plates of the central release control assembly are made of galvanized mild steel in which there is no opening. Therefore, one cannot directly view the interface between the interlock and the cable connection arm. The free end of the interlock lever is housed in a frame, one side of which has a clear plastic window. However, there are no reference marks between this end of the lever and the frame to accurately indicate whether the other end of the interlock is in a blocking position. The manufacturer's operations manual instructs the operator to check whether the free end of the interlock lever is up or down to determine its relative position. With no definitive mark to provide a point of reference, the operator's judgement of up or down becomes subjective and prone to error. The assembly is mounted against the after bulkhead of the lifeboat. One can look into the gap between the two side plates of the assembly housing to get an indication of the position of the interlock lever. However, given the dim lighting inside the lifeboat and that the crew member is wearing a hard hat and a lifejacket, this operation is difficult and uncomfortable and also prone to parallax error. Thus, although the manufacturer's instruction manual indicates the importance of verifying the position of the central release interlock lever, the design and placement of the central release control assembly makes this crucial task difficult. 2.6 The Indication Panel The indication panel displays the status of the lifeboat by using red and green lamps. The panel indicates whether the lifeboat is floating on the water or is suspended in the air and whether it is safe to pull the release handle to open the hooks. Because the two indication panel lamps are multifunctional, each can indicate two separate and contrary operating conditions. In the absence of a straightforward means of seeing the interlock lever's position, the red and green lights are regarded as the most direct and accurate indication of the status of the hook release mechanism. It is essential that the feedback these lights provide be accurate. The lamps' indications are dictated by the contacts made between the cable connection arms, the interlock lever, and their respective microswitches. The positioning of these microswitches (seeFigure4), and the excessive play found in their sensing arms, allowed the panel lamps to falsely indicate the position of the interlock lever. When one or both of the cable connection arms were not being blocked by the interlock lever, the panel lights could incorrectly indicate that they were blocked. 2.7 Typical Interaction Between the Hook Assembly, the Central Release Assembly, and the Indication Panel Once the reset levers have been moved into the fully reset position, their corresponding cable connection arms contact the microswitches, and the green lamp on the panel lights up to indicate that the hooks have been correctly reset. If one or both of the reset levers are now moved down as much as 30 mm, the green lamp will continue to remain lit, thus indicating that the hooks are still reset properly. At some point during the recovery of a lifeboat (when the lifeboat is in the water and the forward and after hooks have been reset correctly), one or both of the suspension rings are inserted into the hooks and the lifeboat rides the waves. While the lifeboat is in the water, the solenoid coil connected to the interlock lever is energized and does not allow the interlock lever to block the cable connection arm. During this same time, the movement of the lifeboat on the waves produces an intermittently applied load on the hook. This load can move the hook release handle down (toward open). This movement, coupled with the associated change in direction of momentum, can allow the reset lever to fall slightly. Its cable connection arm is now lifted and can no longer be blocked by the interlock lever when the lever is released by the solenoid coil. Malfunctioning or 'sticky' electrical contacts in the indication panel or the water pressure switch can also cause the solenoid coil to become energised. These were all tested and found to be working satisfactorily. 2.8 Dynamics of the Accidental Release of the Forward Hook When the reset lever is in its fully reset position, it is 12mm lower than its maximum attainable position. Because the lever's reference marks were never in alignment in the first place, the drop of the reset lever goes unnoticed. Moreover, only the reset lever on the forward hook is readily visible. Static load testing of the hook assembly6 shows that, in this condition, the hook remains fully capable of bearing the applied load of the lifeboat and its contents. However, because the interlock lever is no longer blocking the cable connection arm, the reset lever is now vulnerable to the influence of forces tending to move it downward (open). These forces become predominant when the direction of momentum changes and when the applied load increases. The forces can then cause the reset lever to move down (open). The hook would now open and the suspension ring would be released - just as if the central release handle was pulled. Following the accidental separation of the after hook, the forward reset lever moved unimpeded into the open position because of the increased momentum and load applied at the forward hook when the lifeboat swung about it. The forward suspension ring was thereby released. 2.9 Lifeboat Construction Both of the Pacmonarch's lifeboats complied with all the SOLAS requirements and offered an urgent means of escape for the ship's crew. Nonetheless, a lifeboat may also be required to recover persons in the water. In agitated and confused seas, an unhandy craft, such as a covered lifeboat, is difficult to manoeuvre. Because survivors are normally recovered on the lee side, the provision of access hatches on both sides of the canopy could reduce the time needed to manoeuvre the lifeboat into a position to recover persons in the water and could increase their chance of survival. 2.10 Lifejackets, Seats, and Seatbelts in the Lifeboats The lifeboat was designed for free fall from a height of 3m. Yet when TSB investigators sat in the lifeboat and simulated a fall from about 0.5m, the jar to the spine was considerable when the lifeboat hit the water. The combination of the slope of the lifeboat's canopy, the protrusion of the lifejacket behind the wearers' necks, and the padded headrests forced the wearers to bend their necks forward at a very awkward angle. TSB investigators sat this way for about 30minutes and found it to be quite uncomfortable and painful. Given the cramped quarters and the lack of free space inside the lifeboat, removing the lifejackets and stowing them away would not be easy. People abandoning a vessel may have to spend a considerable length of time in such a vessel, in all kinds of weather. It is difficult to envisage how they could avoid neck injuries while wearing standard lifejackets and bracing themselves against the lifeboat's rolling and pitching. The design of the lifeboat and, in particular, its seating did not make allowance for the fact that the occupants would be wearing standard lifejackets. The lifeboat's uncushioned seating arrangements, combined with the limited headroom when seated and the effect of wearing standard lifejackets, is not conducive to crew comfort and could lead to crew injury. Further, the lifeboat seatbelts are not long enough to efficiently secure a large person to the designated seating when wearing a standard lifejacket. 2.11 Operating Instructions and Training Requirements for the Lifeboat SOLAS regulations address a number of training considerations, including muster lists, emergency instructions, training manuals, and abandon-ship training and drills. This shipboard training helps ensure that the crew will be able to abandon ship successfully in an emergency. Less emphasis is placed on training the crew in lifeboat recovery procedures, presumably because the recovery of the lifeboat is not the primary consideration once the crew has abandoned the vessel. Although all the instruction manuals were written in English, the Ukranian crew of the Pacmonarch had regularly and successfully conducted abandon-ship drills since the vessel was delivered in July. By virtue of this and other continuous proficiency training, before and during their seafaring career, the crew was well versed in launching the lifeboats. The procedure for resetting the hooks, however, was not as well understood by the whole crew. Given the crucial importance of the resetting operation, the availability of an aide-memoire in the lifeboat could have improved the likelihood of the hooks being correctly reset. The instruction manual describes in detail the correct hook-resetting procedure, but this information is not posted inside the lifeboat in written or graphic form for the crew's use. 3.0 Conclusions 3.1 Findings as to Causes and Contributing Factors When the davits swung out as the lifeboat was lowered, the end of the after gripe probably fouled and seized around its bollard. Temporarily hung up on the after gripe, the lifeboat then took a bow-down attitude of about 40, and the after suspension ring moved toward the hook opening. When the gripe freed itself and the load returned to the after falls, the loading most likely caused the after suspension ring to overcome the resistance of the retaining latch and to slip out of the hook. Following the accidental separation of the after hook, the forward reset lever moved unimpeded into the open position because of the increased momentum and load applied at the forward hook when the lifeboat swung about it. The forward suspension ring was thereby released. 3.2 Findings as to Risk Although the instruction manual describes in detail the correct hook-resetting procedure, this information is not posted inside the lifeboat in written or graphic form for the crew's use. The after hook's aft-facing orientation places its reset lever where the operator cannot easily manipulate it or see that the lever has remained in place after being reset. The misaligned hook reset reference marks did not provide the crew with an accurate indication that the reset lever was in the fully up position and, critically, that the hooks were properly reset. The manufacturer's operations manual instructs the operator to check whether the free end of the central release interlock lever is up or down to determine its relative position. With no definitive mark to provide a point of reference, an operator's judgement of up or down becomes subjective and prone to error. Although the manufacturer's instruction manual indicates the importance of verifying the position of the central release interlock lever, the design and placement of the central release assembly makes this crucial task difficult. Because the two indication panel lamps are multifunctional, each can indicate two separate and contrary operating conditions. When one or both of the cable connection arms were not being blocked by the interlock lever, the indication panel lights could incorrectly indicate that they were blocked. The manufactured length from the end of the after hook's load-transmitting surface to the circumference of its pivot was 121.8mm. Because the equipment was new and the wear minimal, this length is a deviation from the designed profile and indicates a manufacturing nonconformity. The turning moment on the hook reset lever indicates that the profile of the lever's load-bearing cam was not of a uniform radius, concentric with the centre of rotation. Because the equipment was new and wear was undetected and unlikely, this profile is a deviation from the designed profile and indicates a manufacturing nonconformity. During recovery of the lifeboat from the water, the movement of the lifeboat on the waves produces an intermittently applied load on the hook. This load can move the hook release handle down (toward open). 3.3 Other Findings The design of the lifeboat and, in particular, its seating did not make allowance for the fact that the occupants would be wearing standard lifejackets. The design of the lifeboat's uncushioned seating arrangements, combined with the limited headroom when seated and the effect of wearing standard lifejackets, is not conducive to crew comfort and could lead to crew injury. The lifeboat seatbelts are not long enough to efficiently secure a large person to the designated seating when wearing a standard lifejacket. The provision of access hatches on both sides of the lifeboat canopy could reduce the time to manoeuvre the lifeboat into a position to recover persons in the water and could increase their chance of survival. Although the lifeboat was designed for free fall from a height of 3m, a simulated fall during testing from about 0.5m caused considerable jarring and discomfort to the occupants when the lifeboat hit the water. 4.0 Safety Action 4.1 Action Taken The TSB sent a Marine Safety Advisory (MSA 05/01) to the Bahamas Maritime Authority, with a copy to the vessel owner, and a Marine Safety Information Letter (MSI 02/01) to the International Association of Classification Societies, the Bahamian flag state, Transport Canada, Class NK, the owners of the Pacmonarch, the manufacturers of the lifeboat and davits, and the ship builder, to advise them of the ergonomic and design deficiencies of the hook and central release control assemblies. The MSI also stressed the importance of ensuring that the interlock lever blocks the cable connection arms. In response, Transport Canada Marine Safety (TCMS) issued Port State Control Instruction No3, based on this and other similar accidents, concerning the premature release of lifeboat safety hooks. Inspectors are referred to Ship Safety Bulletin 05/2000 and other documents to ensure that, during inspections, all launching systems are inspected and that the crew is familiar with the correct sequences for launching and retrieving survival craft. If deficiencies are detected, the vessel is to be detained. TCMS also advised that Australia, Canada, and New Zealand submitted a joint paper (DE 45/17) to the International Maritime Organization's Subcommittee on Ship Design and Equipment, on 14December2001, to raise awareness of the safety issues surrounding the operation and maintenance of totally enclosed lifeboats. The paper was intended to assist the subcommittee in its consideration of accident prevention and set out various possible short, medium and long term measures. Three tanker operator organizations (OCIMF, INTERTANKO SIGTTO), in supporting these papers, submitted further information to the subcommittee on 25January2002. This information included additional items about lifeboat design, operations, standards, guidelines, training, and maintenance aspects for consideration in the short and medium terms. The subcommittee met in March2002 and will next meet in 2003. The TSB convened a meeting at its Engineering Laboratory facility in April2001. Representatives of the lifeboat, davit, and ship manufacturers, the vessel's owner, and the lifeboat manufacturer's sales agent attended. Deficiencies noted in the design of the hook and release mechanisms were discussed and demonstrated. The discussion was thorough and constructive, and particular emphasis was placed on the design of the user-machine interface. After the meeting, Lasco Shipping Co., the operating managers of the vessel, decided to replace all four lifeboats on the Pacmonarch and the Pacemperor with lifeboats of another design made by a different manufacturer. The lifeboat manufacturer, Nishi-F Co. Ltd., responded to the MSI with more information regarding the reset position and process for the hooks, the system of lights and the interlocks. After an investigation, the Bahamas Maritime Authority has recommended that the classification society, Class NK, reconsider approval of this type of lifeboat until the manufacturers make changes in accordance with its recommendations. These apply to all lifeboats, both existing and new, and are, inter alia: Manufacturers to consider making a viewing port on the side of the central release assembly, to allow a direct view that the interlock lever is actually blocking the cable connection arm, when the hook mechanism has been reset. Manufacturers to consider fitting the aft hook with the open end facing forward, so that the red alignment marks are easily viewed when being reset by the lifeboat crew. Verify that the red alignment marks align correctly before delivery. The Japan Classification Society, Class NK, responded to the MSI, advising that some 340boats, manufactured by Nishi-F Co. Ltd., had been installed on ships using their services since 1987. Class NK sent an advisory note to all their offices and to all known owners of vessels with this type of lifeboat. Class NK has had discussions with the lifeboat manufacturer, as a result of which they have decided to : cease all further production of this design of hook release mechanism; deliver safety bars to all other existing ships with lifeboats of this type. These safety bars will be inserted into the central release assembly, to prevent inadvertent, premature opening of the hooks; simplify their instruction and operation manual so that it can be more easily understood; and develop training videos for the operation and deployment of the lifeboat. The International Association of Classification Societies (IACS) has circulated the information in the MSI to all their members who will consider what action is deemed necessary. 4.2 Safety Concern Lifeboat Ergonomics The advisory and information letters mentioned in Section 4.1 dealt with the practicalities of lifeboat safety hooks, their design and the ergonomics involved in releasing or re-attaching the blocks. It is acknowledged that officials and organizations have taken action and that the problems encountered have been brought to the attention of IMO. However, the Board remains concerned that injury to crew may result during lifeboat drill or the abandonment of the vessel. Tests conducted by TSB personnel in the Pacmonarch's davit launched enclosed lifeboat indicated that the design of the lifeboat's seating arrangements did not allow crew, who were wearing standard lifejackets, to be safely seated. The lifeboat was designed to fall from a height of 3m above the water but, during a test from about 0.5m, the occupants suffered considerable jarring when the lifeboat hit the water. The uncushioned seating, limited headroom, too short seat belts and the seated position of a crew person wearing a standard lifejacket create the conditions which may, as a result of the boat being released, cause personal injury.